Hi all.
In the process of building up the randomHR QGIS plugin[1], we stumbled
across a more general question: which is the most appropriate metric for
measuring overlap among HRs?
Should it be:

1) area of overlapping / area occupied by at least one HR
2) area of overlapping / sum of areas of all HRs
3) N x area of overlapping / sum of areas of all HRs
      (where N = number of overlapped HRs in given place)
4) something else

In the attached file, let's assume each square is 1ha big, so each range
is 2ha
big.

example A:
method 1)  1 / 3 = 33%
method 2)  1 / 4 = 25%
method 3)  2 / 4 = 50%

example B:
method 1)  1 / 4 = 25%
method 2)  1 / 6 = 17%
method 3)  3 / 6 = 50%

Opinions and suggestions welcome.
All the best.
-- 
Paolo Cavallini, see: * http://www.faunalia.it/pc *

[1] https://www.faunalia.it/dokuwiki/doku.php/public/animove_qgis_plugins

<<inline: overlap.png>>

_______________________________________________
AniMov mailing list
[email protected]
http://www.faunalia.com/cgi-bin/mailman/listinfo/animov

Reply via email to